457 ARTICLE

Patterns of genome size diversity in bats (order Chiroptera)1 Jillian D.L. Smith, John W. Bickham, and T. Ryan Gregory

Abstract: Despite being a group of particular interest in considering relationships between genome size and metabolic param- eters, bats have not been well studied from this perspective. This study presents new estimates for 121 “microbat” species from 12 families and complements a previous study on members of the family Pteropodidae (“megabats”). The results confirm that diversity in genome size in bats is very limited even compared with other mammals, varying approximately 2-fold from 1.63 pg in Lophostoma carrikeri to 3.17 pg in Rhinopoma hardwickii and averaging only 2.35 pg ± 0.02 SE (versus 3.5 pg overall for mammals). However, contrary to some other vertebrate groups, and perhaps owing to the narrow range observed, genome size correlations were not apparent with any chromosomal, physiological, flight-related, developmental, or ecological characteristics within the order Chiroptera. Genome size is positively correlated with measures of body size in bats, though the strength of the relation- ships differs between pteropodids (“megabats”) and nonpteropodids (“microbats”).

Key words: Chiroptera, genome size, C-value, flight, metabolism. Résumé : Bien qu’elles constituent un groupe présentant un intérêt particulier pour l’étude des relations entre la taille du génome et les paramètres métaboliques, les chauves-souris n’ont pas été bien étudiées sous cet angle. Dans ce travail, les auteurs présentent des estimés pour 121 espèces de “microchiroptères” appartenant a` 12 familles et ceci vient compléter une étude antérieure sur des membres de la famille des Pteropodidae (“mégachiroptères”). Les résultats confirment que la diversité de la taille des génomes chez les chauves-souris est limitée par rapport a` d’autres mammifères, variant environ du simple au double entre 1,63 pg chez le Lophostoma carrikeri a` 3,17 pg chez le Rhinopoma hardwickii pour une moyenne de 2,35 pg ± 0,02 (contre 3,5 pg globalement chez les mammifères). Cependant, contrairement a` ce qui est observé chez d’autres vertébrés, peut-être en raison de la faible variation observée, la taille du génome ne semblait pas corrélée avec des caractéristiques chromosomiques, physio- logiques, liées au vol, développementales ou écologiques chez les chiroptères. La taille du génome était cependant positivement corrélée avec divers paramètres de la taille du corps chez les chauves-souris, bien que le degré de corrélation différait chez les ptéropodidés (“mégachiroptères”) et les non-ptéropodidés (“microchiroptères”). [Traduit par la Rédaction]

Mots-clés : Chiroptera, taille du génome, valeur C, vol, métabolisme.

For personal use only. Introduction 132.83 pg), especially with the objective of identifying phenotypic The order Chiroptera is the second largest in mammals after correlates that could help to explain this astounding diversity rodents, and with >1100 species in 18 families it represents more (Gregory 2013). However, increasing attention has been paid more than one-fifth of all mammalian diversity (Wilson and Reeder recently to explaining the apparent constraint within some 2005). While bats share many commonalities with other mam- otherwise diverse vertebrate taxa, in particular among birds mals, they are the only mammals capable of true flight and, along (Andrews et al. 2009). with birds and the extinct pterosaurs, one of only three vertebrate It has been hypothesized that genome size may be constrained in taxa to have evolved this highly specialized mode of locomotion. some vertebrate groups with high metabolic rates. This is particu- Despite their diversity, abundance, and unique biology, bats re- larly relevant in those taxa subject to the extreme metabolic de- main poorly studied from several important perspectives. Nota- mands of powered flight owing to the relationship between genome bly, research into the diversity of genome size in bats is one such size, cell size, and mass-specific metabolic rate (Hughes and Hughes area that is lacking. Prior to recent work by the current authors 1995; Gregory 2002a; Organ and Shedlock 2009). Specifically, smaller Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 (J.D.L. Smith and Gregory 2009; present study), estimates available cells—which are associated with small genomes—have a higher sur- in the Animal Genome Size Database covered only 62 species from face area to volume ratio, allowing for improved gas exchange to 7 families (Gregory 2013). meet metabolic demands (Szarski 1983; Gregory 2001b). In one analysis, Hughes (1999) used a comparatively small data- Genome size constraint set for birds to suggest that strong flyers have smaller genomes The genome sizes of animals as a whole range more than 7000-fold. than weak flyers or flightless taxa (see also Gregory (2005a)). More Even among vertebrates, genome sizes vary by a factor of more recently, Andrews et al. (2009) used a much larger, newly gener- than 350. Not surprisingly, many studies on genome size in ani- ated dataset to examine the relationship between genome size mals have focused on groups with large amounts of variability, and flight using specific wing parameters. In this case, genome such as amphibians (1C = 0.95 – 120.60 pg) and “fishes” (1C = 0.35 – size was positively correlated with wing loading index (an inverse

Received 8 March 2013. Accepted 8 July 2013. Corresponding Editor: Jillian Bainard. J.D.L. Smith and T.R. Gregory. Department of Integrative Biology, University of Guelph, 50 Stone Road E., Guelph, ON N1G 2W1, Canada. J.W. Bickham. Battelle Memorial Institute, 10777 Westheimer, Suite 1105, Houston, TX 77042, USA. Corresponding author: T. Ryan Gregory (e-mail: [email protected]). 1This article is one of a selection of papers published in this Special Issue on Genome Size Evolution.

Genome 56: 457–472 (2013) dx.doi.org/10.1139/gen-2013-0046 Published at www.nrcresearchpress.com/gen on 23 September 2013. 458 Genome Vol. 56, 2013

indicator of flight efficiency) in passerine birds, lending support velopmental traits have also been studied in fishes, birds, and to the flight constraint hypothesis (Andrews et al. 2009). Fossil mammals (including within primates and rodents), although re- evidence also shows that avian dinosaurs and pterosaurs had ports have been conflicting with no clear indication of its impor- small genomes relative to nonflying lineages (Organ et al. 2007; tance to genome size (Morand and Ricklefs 2001, 2005; Gregory Organ and Shedlock 2009). In light of these studies on other flying 2002c; E.M. Smith and Gregory 2009). vertebrates, it seems that flight may indeed impose a constraint It might be tempting to assume that because links have been on genome size. Clearly, a study of bats also represents a critical found between genome size and one or more adaptive characters, component of this line of inquiry. that this alone may be responsible for observed diversity in DNA amount among species. However, there are nonadaptive possibil- Explaining genome size diversity within bats ities that should also be considered. For example, it is possible While genome size in bats appears constrained based on the lim- that mutational processes at the level of whole chromosomes play ited sampling conducted to date, little is known about the causes and an important role. Indeed, a link has been found between genome consequences of this variation. Flight is a proposed constraint on size and chromosomal number in teleost fishes (Mank and Avise genome size; however, previous studies have not considered rela- 2006; E.M. Smith and Gregory 2009). Whether genome size is driven tionships between genome size and flight or any other organism by chromosomal duplications or losses, or whether more DNA is level traits within Chiroptera, such that work on this group lags associated with higher rates of breakage, remains an open issue. It is, behind that focusing on other vertebrates (reviewed in Gregory therefore, worth examining these associations in bats and other 2005a). Mammals, like birds, demonstrate strong positive links be- groups. tween genome size and cell size and inverse relationships between Another intriguing possibility relates genome size to mutation genome size and metabolic rate (Vinogradov 1995; Gregory 2000). In rate, genetic drift, and population size. From the perspective of some cases, relationships can be identified within individual orders, population genetics, smaller populations are more susceptible to such as a correlation between genome size and body size in rodents the influence of genetic drift, and thus the genomes of these (Gregory 2002c) or between genome size, cell size, and flight param- animals are more likely to retain mildly deleterious mutations. eters within the avian order Passeriformes (Andrews et al. 2009). Lynch and Conery (2003) proposed that animals have larger ge- It, therefore, remains an important question whether cell- and nomes relative to prokaryotes and unicellular organisms because organism-level traits, especially those relevant to flight, are related to their smaller population sizes passively allowed for the accumu- genome size diversity among bats; and whether this can help to lation of gene duplications and mobile genetic elements. While explain patterns within this highly diverse mammalian order, in empirical support for this model has been limited (Gregory and addition to assessing the larger issue of flight-related constraints Witt 2008), this remains a question worthy of analysis. Though across bats generally. population size estimates are very difficult to obtain, body size While bats and birds share the feature of powered flight and has been suggested as a suitable proxy (Lynch and Conery 2003). high relative metabolisms, mammals differ from other vertebrates in that they possess enucleated red blood cells. This expulsion of Taxonomy nuclei during red blood cell development results in exceptionally Bats are interesting targets for genome size study owing to their compact erythrocytes that allow for efficient gas exchange without incredible diversity and apparent genomic contraint, but they necessarily requiring extremely reduced genome sizes (Gregory also present an interesting case because of ongoing controversey 2000). This further emphasizes the need to study questions relating over phylogenetic relationships within order Chiroptera. Early

For personal use only. to genome size constraints in bats, because this critical difference classifications grouped all bats together based on their flight abil- between birds and mammals could decouple, or at least weaken, the ity, with later separation of bats into two groups, the Megachirop- relationship between genome size and metabolism. tera (family Pteropodidae) and the Microchiroptera (all other Of course, flight is not the only characteristic that could be respon- bats). This division was initially based largely on overall body size; sible for influencing genome size in birds and bats. The link between however, with time many additonal morphological and ecological genome size, cell division, and cell size can lead to associations with differences supporting the separation were found between the other biological features, and it has been suggested that organ com- two groups. Monophyly was traditionally assumed but was chal- plexity, development, and ecological lifestyle have been important lenged in the 1980s with the hypothesis that megabats were more in the evolution of genome size among other vertebrates (e.g., closely related to primates, and thus that flight evolved twice within Gregory 2002c, 2005a; Andrews and Gregory 2009). For example, mammals (Pettigrew 1986). Cladistic analysis of morphological char- Roth et al. (1994) discovered that amphibians exhibit a negative cor- acteristics appeared to support this hypothesis; however, since relation between genome size and brain complexity because fewer, the application of molecular phylogenetics, the monophyly of

Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 less well differentiated neurons can be fit within the braincase when bats has once again become widely accepted (Bailey et al. 1992). the cells are larger and divide more slowly. This is important, as brain Nonetheless, debate persists regarding the validity of the complexity and function are highly associated with many aspects of megabat–microbat division. Some recent classifications have animal behaviour and ecology. Similarly, a negative relationship ex- not supported this distinction, instead dividing bats into the ists between genome size and relative brain size in parrots (Andrews Yangochiroptera and Yinpterochiroptera. In this case, “mega- and Gregory 2009). By extension, it is possible that genome size di- bats” are nested within the “microbat” families, being more versity is associated with feeding, habitat, or social behaviour as closely related to the families Rhinolophidae, Megadermatidae, these can be influenced by cognitive capabilities (Andrews and and Rhinopomatidae in the Yinpterochioptera (Teeling et al. Gregory 2009). In bats, brain size has been linked to foraging ecology, 2002). By contrast, a recent study of the prestin gene, which is feeding type, and habitat complexity in a number of studies (e.g., involved in hearing and suggested to be very important for Eisenberg and Wilson 1978; Safi and Dechmann 2005; Ratcliffe et al. echolocation, has provided support for the traditional mega- 2006). bat–microbat division; although this evidence has been criti- A negative relationship between genome size and cell division rate cized based on the nature of the gene (Li et al. 2008). At present, has long been observed in many animal groups as well as in plants this issue remains unresolved. However, it is clear that the (e.g., Van’t Hof and Sparrow 1963; Grosset and Odartchenko 1975a, Pteropodidae are morphologically, physiologically, and ecolog- 1975b). This association suggests that developmental traits may also ically unique among the bats, and as such the present study be linked to genome size at the whole organism level. In fact, an includes analyses across all bats as well as within pteropodids inverse relationship has been found between genome size and devel- (“megabats”) and nonpteropodids (“microbats”) treated as opmental rate in amphibians (Bachmann 1972b; Gregory 2003). De- functionally (if not phylogenetically) distinct categories.

Published by NRC Research Press Smith et al. 459

Questions and predictions Genome size data There are many intriguing questions to be asked regarding ge- In total, 650 samples from 448 individuals representing 121 spe- nome size diversity in vertebrates in general, of which biologically cies and 12 families were obtained from the Lubee Bat Conser- unique but previously overlooked groups like bats represent a vancy (Gainesville, Florida), the Royal Ontario Museum (Toronto, particularly interesting case. In an effort to fill this gap and to Ontario), the University of Alaska Museum of the North (Fair- shed light on the evolution of bats and their genomic character- banks, Alaska), and the collection of one of the authors (J.W.B., istics, this study poses and attempts to answer the following spe- Purdue University). Samples contributed by the Lubee Bat Conser- cific research questions: vancy consisted of air-dried blood smears taken from captive bats during routine veterinary care. Samples from Purdue University, (1) What is the extent of genome size diversity in bats, and is this the Royal Ontario Museum, and the Museum of the North were universally constrained relative to other mammals? prepared on site from frozen kidney, liver, and (or) spleen tissues If flight constrains genome size in vertebrates, then all new stored at –80 °C with no added preservatives or media. estimates for bats are predicted to be small relative to nonfly- Genome size (GS) estimation was conducted using Feulgen im- ing mammals. age analysis densitometry (FIA), following best practice methods (2) Are cytogenetic features related to genome size diversity as described in detail by Hardie et al. (2002). Briefly, slides were among bats? fixed overnight at room temperature in a methanol–formalin– If chromosome-level processes are responsible for the di- glacial acetic acid solution of 85:10:5 (by volume), followed bya2h versity in genome size observed among bats, then a relation- hydrolysis in 5.0 N HCl and subsequent staining for2hinSchiff ship between chromosome number and genome size is reagent; a series of bisulphate and water rinses was used to termi- expected to exist. nate staining. A minimum of 50 nuclei was measured per sample (3) Are differences in genome size among bats linked to body using the Bioquant Life Science version 8.00.20 software package size? and an Optronics DEI-750 CE three-chip CCD camera mounted on If cell size, and not only cell number, affects body size in a Leica DM LS microscope at 100× magnification. The resulting bats then genome size is predicted to be positively correlated integrated optical densities (IODs) were converted to genome size with body size among bats as it is among rodents and birds. in picograms using the IODs of two standards of known genome Alternatively (or in addition), body size may represent a proxy size: Sus scrofa (2.91 pg) and Bos taurus (3.56 pg) (Gregory 2013). for population size, which has been proposed to correlate Standards were stained in tandem and were of the same tissue with genome size. This latter prediction can be tested by in- type and preservation method (fresh or frozen) as the samples. vestigating other measures of population size, such as roost Both standards were used for all estimates; averages of the two size. estimates gave the final genome size estimate. The additional (4) Can constraints related to flight explain the observed diver- biological parameters compared with these genome size data are sity of genome size within bats? summarized below (Smith 2009). If flight constraints determine patterns of genome size di- versity within bats, and not just relative to other mammals, Chromosomal data then bats are predicted to have small cells and to exhibit • The diploid number of chromosomes. correlations between genome size and metabolic rate and • The fundamental number of chromosome arms, an indicator of wing parameters related to flight ability (e.g., wing loading). chromosome structure; calculation assumes metacentric, sub- For personal use only. (5) Are neurological constraints relevant to genome size diver- metacentric, and subtelocentric chromosomes have two arms, sity among bats? while telocentric or acrocentric have one arm. If the metabolic expense of the brain influences genome • The ratio of diploid chromosome number to the fundamental size diversity (due to investment in brain tissue, and (or) con- number of chromosome arms. straints on brain complexity based on neuron size vs. num- ber), then bats should display relationships between genome Cell size data size and indicators of investment in brain tissue (e.g., relative • ␮ brain volume). Mean dry cell diameter of erythrocytes in micrometers ( m). • Mean corpuscular volume in femtoliters (fL): calculated by divid- (6) Are developmental parameters related to diversity in genome ing the hematocrit (L) by the red blood cell count (millions/␮L). size among bats? If genome size is linked to cell division in bats, then this Body size data may be predicted to produce relationships between genome Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 size and parameters related to reproduction and develop- • Body mass in grams (g): recorded primarily from Smith et al. ment, such as gestation time or number of litters per year. (2003) with missing values obtained from alternate sources (7) Is diversity in genome size among bats associated with any when available. ecological features? • Head and body length in millimeters (mm) (distance from the If genome size is constrained by (or influences) body size, tip of the nose to the base of the tail). metabolism, cognition, and (or) development, then it might be expected that genome size will be related to ecological Physiological data features such as feeding, habitat, or sociality. • Absolute basal metabolic rate (BMR) in milliliters of oxygen per

hour (ml.O2/hr). Materials and methods • Body mass (g): recorded from the individuals used to measure The dataset used in this study is composed of novel genome size basal metabolic rate. estimates for nonpteropodid bats (“microbats”) (Table 1), 43 esti- • Relative basal metabolic rate (RBMR) in milliliters of oxygen per

mates for pteropodid bats (“megabat”) produced by the same hour per gram of body mass (ml.O2/hr.g): calculated by dividing methodology (J.D.L. Smith and Gregory 2009), and a series of bio- the absolute basal metabolic rate by the mass of the individuals logical parameters collected from the literature2 (Smith 2009). measured.

2Supplementary data are available with the article through the journal Web site at http://nrcresearchpress.com/doi/suppl/10.1139/gen-2013-0046.

Published by NRC Research Press 460 Genome Vol. 56, 2013

Table 1. New mean haploid genome size estimates (GS, in pg) for chiropteran species, including the standard error for estimates obtained from more than one individual. Tissue Tissue Family Species Common Name GS (pg) SE n (F/M/U) type source Emballonuridae 2.67 0.49 2 Saccopteryx bilineata Greater sac-winged bat 3.16 1F SP 1 Taphozous sp. Tomb bat sp. 2.18 0.03 2U KC 1 Hipposideridae 2.57 0.07 4 Hipposideros caffer Sundevall's leaf-nosed bat 2.44 0.05 2F, 2M KC, SP 1 Hipposideros commersoni Commerson's leaf-nosed bat 2.48 0.10 2M KC 1 Hipposideros cyclops Cyclops leaf-nosed bat 2.72 1M KC 1 Hipposideros ruber Noack's leaf-nosed bat 2.65 1F KC 1 Megadermatidae 2.54 0.19 2 Cardioderma cor Heart-nosed bat 2.72 0.09 2F, 1M KC 1 Lavia frons Yellow-winged bat 2.35 1M KC 1 Molossidae 2.53 0.08 6 Chaerephon nigeriae Nigerian free-tailed bat 2.47 0.06 1F, 1M KC 1 Chaerephon pumilus Little free-tailed bat 2.40 1M KC 1 Eumops perotis Greater bonneted bat 2.84 1F LV 1 Molossus molossus Pallas's mastiff bat 2.41 1F KC 1 Molossus rufus Black mastiff bat 2.72 1F SP 1 Sauromys petrophilus Roberts's flat-headed bat 2.32 0.15 1M, 1U KC, LV 1 Mormoopidae 2.45 0.10 5 Mormoops megalophylla Peters's ghost-faced bat 2.83 1M SP 1 Pteronotus davyi Davy's naked-backed bat 2.28 0.06 1F, 1M SP 1 Pteronotus parnellii Common mustached bat 2.35 0.04 3F, 2M LV, SP 1 Pteronotus personatus Wagner's mustached bat 2.43 0.00 2M SP 1 Pteronotus personatus Wagner's mustached bat 2.35 1F SP 1 psilotis Natalidae 2.30 1 Natalus sp. Greater funnel-eared bat 2.30 1M KC 1 Noctilionidae 2.33 0.05 2 Noctilio albiventris Lesser bulldog bat 2.28 1U KC 1 Noctilio leporinus Greater bulldog bat 2.37 1F SP 1 Nycteridae 2.77 0.05 3 Nycteris arge Bates's slit-faced bat 2.83 1M KC 1 Nycteris hispida Hairy slit-faced bat 2.81 1F KC 1 Nycteris major Dja slit-faced bat 2.66 0.09 2M KC 1

For personal use only. Phyllostomidae 2.43 0.04 34 Artibeus cinereus Gervais's fruit-eating bat 2.44 0.03 2M KC 1 Artibeus intermedius Intermediate fruit-eating bat 2.43 0.03 1F, 2M SP 1 Artibeus hirsutus Hairy fruit-eating bat 2.28 0.07 2U KC, LV 1 Artibeus jamaicensis Jamaican fruit-eating bat 2.65 0.07 3F LK, SP 1, 2 Artibeus lituratus Great fruit-eating bat 2.57 0.02 3F, 5M LK, LV, SP 1, 2, 3 Artibeus obscurus Dark fruit-eating bat 2.44 0.01 26U KC, LV 4 Carollia brevicauda Silky short-tailed bat 2.48 0.04 1F, 1M LV 1 Carollia castanea Chestnut short-tailed bat 2.54 1F LV 1 Carollia perspicillata Seba's short-tailed bat 2.63 0.01 30U KC, LV 1, 3, 4 Centurio senex Wrinkle-faced bat 2.73 1F SP 1 Chiroderma villosum Hairy big-eyed bat 2.67 0.01 2F SP 1 Choeronycteris mexicana Mexican long-tongued bat 2.35 1U SP 1 Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 Chrotopterus auritus Woolly false vampire bat 2.53 0.10 2F, 2M LV, SP 1, 3 Desmodus rotundus Common vampire bat 2.39 0.11 2F, 1U SP 1 Glossophaga longirostris Miller's long-tongued bat 2.22 0.06 2M KC, LV 1 Glossophaga soricina Pallas's long-tongued bat 2.34 0.04 2F, 1M LV, SP 1 Leptonycteris nivalis Mexican long-nosed bat 2.36 1M SP 1 Lichonycteris obscura Dark long-tongued bat 2.72 1F KC 1 Lophostoma carrikeri Carriker's round-eared bat 1.63 1M KC, LV 1 Macrotus waterhousii Waterhouse's leaf-nosed bat 2.67 1M SP 1 Micronycteris hirsuta Hairy big-eared bat 2.24 0.04 1F, 2M KC 1 Mimon cozumelae Cozumelan golden bat 2.30 0.04 2F, 2M SP 1 Mimon crenulatum Striped hairy-nosed bat 2.19 0.04 1F, 1M KC, LV 1 Phylloderma stenops Pale-faced bat 2.22 0.03 2M KC 1 Phyllostomus discolor Pale spear-nosed bat 2.58 1F LV 1 Platyrrhinus helleri Heller's broad-nosed bat 2.63 1U KC 1 Platyrrhinus lineatus White-lined broad-nosed bat 2.43 0.05 2F, 2M LV 3 Pygoderma bilabiatum Ipanema broad-nosed bat 2.47 0.10 2M LV 3 Sturnira lilium Little yellow-shouldered bat 2.61 0.08 3F, 3M LV, SP 1, 3

Published by NRC Research Press Smith et al. 461

Table 1 (continued). Tissue Tissue Family Species Common Name GS (pg) SE n (F/M/U) type source Sturnira tildae Tilda's yellow-shouldered bat 2.53 1M LV 1 Tonatia bidens Greater round-eared bat 2.27 0.06 1F, 1M KC, LV 1 Trachops cirrhosus Fringe-lipped bat 2.22 0.04 1F, 1M KC, LV 1 Trinycteris nicefori Niceforo's bat 2.32 1M KC 1 Uroderma bilobatum Common tent-making bat 2.42 0.08 1F, 1U KC, SP 1 Rhinolophidae 2.30 0.04 2 Rhinolophus darlingi Darling's horseshoe bat 2.33 0.04 1F, 3U KC, LV 1 Rhinolophus landeri Lander's horseshoe bat 2.26 1M LV 1 Rhinopomatidae 3.17 1 Rhinopoma hardwickii Lesser mouse-tailed bat 3.17 1M KY 1 Vespertilionidae 2.33 0.02 59 Antrozous pallidus Pallid bat 2.67 0.05 1F, 3M KC, LV 1 Bauerus dubiaquercus Van Gelder's bat 2.62 0.14 1M, 1U LV, SP 1 Chalinolobus gouldii Gould's wattled bat 2.29 1F KC 1 Chalinolobus morio Chocolate wattled bat 2.45 0.09 3F, 2M KC, LV 1 Eptesicus bottae Botta's serotine 2.46 0.02 2F, 1M KC, LV 1 Eptesicus brasiliensis Brazilian brown bat 2.36 1F SP 1 Eptesicus furinalis Argentinian brown bat 2.34 0.03 2F, 2M LV 1 Eptesicus fuscus Big brown bat 2.32 0.04 2F, 1M LV 1 Eptesicus hottentotus Long-tailed serotine 2.37 0.04 2F, 2M KC 1 Eptesicus serotinus Common serotine 2.38 0.00 2M KC, LV 1 Glauconycteris beatrix Beatrix butterfly bat 2.13 1F KC 1 Harpiocephalus harpia Lesser hairy-winged bat 2.41 1M KC 1 Hesperoptenus blanfordi Blanford's bat 2.54 1F SP 1 Hesperoptenus tickelli Tickell's bat 2.83 0.05 1F, 1U KC, SP 1 Histiotus macrotus Big-eared brown bat 2.37 1F KC, LV 1 Kerivoula argentata Damara woolly bat 2.30 1M KC 1 Laephotis botswanae Botswanan long-eared bat 2.19 1M KC, LV 1 Laephotis namibensis Namibian long-eared bat 2.28 0.04 5F, 1M KC, LV 1 Laephotis wintoni De Winton's long-eared bat 2.03 1M KC 1 Lasionycteris noctivagans Silver-haired bat 2.22 1F LV 3 Lasiurus ega Southern yellow bat 2.46 0.05 2F, 3M KC, LV 1 Lasiurus insularis Cuban yellow bat 2.54 1U LV, SP 1 Lasiurus intermedius Northern yellow bat 2.43 0.05 2F, 2M KC 1 Lasiurus xanthinus Western yellow bat 2.37 0.07 1F, 2M KC 1

For personal use only. Lasiurus borealis Eastern red bat 2.31 0.06 2F, 1M KC, LV 1 Lasiurus cinereus Hoary bat 2.37 0.06 1F, 1M LV 1 Lasiurus seminolus Seminole bat 2.46 0.04 1F, 2U KC, LV 1 Miniopterus fraterculus Lesser long-fingered bat 1.91 0.00 1F, 1M KC, LV 1 Miniopterus inflatus Greater long-fingered bat 2.04 0.05 1M, 2U KC, LV 1 Miniopterus schreibersii Schreibers's long-fingered bat 1.93 0.00 1F, 2M KC 1 Murina florium Flores tube-nosed bat 2.00 1M LV 1 Myotis bocagii Rufous myotis 2.35 0.13 2F LV 1 Myotis keenii Keen's myotis 2.21 0.04 3M KC, LV 1, 3 Myotis lucifugus Little brown myotis 2.26 0.07 2F, 4M KC, LV 1, 3 Myotis riparius Riparian myotis 2.20 1U KC 1 Myotis septentrionalis Northern myotis 2.20 0.07 2F, 2M KC 1 Myotis velifer Cave myotis 2.30 0.07 2F, 2M KC, LV 1 Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 Myotis yumanensis Yuma myotis 2.30 0.02 3F KC, LV 1 Neoromicia capensis Cape serotine 2.14 0.05 2F, 2M KC 1 Neoromicia guineensis Guinean serotine 2.61 1F KC 1 Neoromicia somalicus Somali serotine 2.19 0.05 2F, 2M KC, LV 1 Neoromicia helios Samburu pipistrelle 2.26 0.06 1F, 2M KC, LV 1 Neoromicia nanus Banana pipistrelle 2.26 0.11 1F, 2M KC 1 Nycticeinops schlieffeni Schlieffen's twilight bat 2.32 0.14 2U KC 1 Nycticeius humeralis Evening bat 2.19 0.01 2F, 1M, 1U KC, LV 1 Nyctophilus gouldi Gould's long-eared bat 2.34 0.04 1F, 2M LV 1 Otonycteris hemprichii Hemprich's desert bat 2.47 1U KC 1 Pipistrellus subflavus Eastern pipistrelle 2.61 0.05 2F, 1M KC 1 Pipistrellus coromandra Indian pipistrelle 1.99 0.03 2F, 2M LV 1 Pipistrellus hesperus Western pipistrelle 2.29 1U KC 1 Pipistrellus kuhlii Kuhl's pipistrelle 2.04 0.05 1F, 1M KC 1 Pipistrellus rusticus Rusty pipistrelle 2.31 1M LV 1 Plecotus sp. Long-eared bat 2.23 1F KC 1 Scotoecus hindei Hinde's lesser house bat 2.41 0.07 1F, 3M LV 1

Published by NRC Research Press 462 Genome Vol. 56, 2013

Table 1 (concluded). Tissue Tissue Family Species Common Name GS (pg) SE n (F/M/U) type source Scotophilus nigrita Giant house bat 2.41 1M LV 1 Scotophilus nux Nut-colored house bat 2.54 0.06 2F KC, LV 1 Scotophilus viridis Green house bat 2.56 0.05 2F, 2M KC, LV 1 Vespadelus darlingtoni Large forest bat 2.43 0.04 2F, 1M KC, LV 1 Vespadelus regulus Southern forest bat 2.42 0.05 1F, 2M LV 1 Note: Genome size estimates were measured using Feulgen image analysis densitometry with at least 50 nuclei measured per sample with two samples being measured from most individuals. The number of individuals sampled is given as the number of females, males, or individuals of unknown sex (n (F/M/U)). The tissue type used for measurements is indicated as kidney (KC), leukocytes (LK), liver (LV), or spleen (SP). Standards were the same tissue type from Bos taurus (GS = 3.56 pg) and Sus scrofa (GS = 2.91 pg) for all samples. Chiropteran tissues were collected from 1: John Bickham, Purdue University, West Lafayette, Indiana; 2: Lubee bat Conservancy, Gainesville, Florida; 3: University of Alaska Museum of the North, Fairbanks, Alaska; 4: Royal Ontario Museum, Toronto, Ontario.

• Body temperature in degrees Celsius (°C): taken from animals 3: Omnivore at rest at their usual ambient temperature. 4: Carnivore/Sanguivore • Geographic distribution: bats are divided categorically into bio- Flight data geographic regions: The defining study on bat flight and wing parameters by 1: Nearctic Norberg and Rayner (1987) was used for most wing parameters. 2: Palearctic When necessary, corrections were made to additional values us- 3: Neotropic ing the methods described by Norberg and Rayner (1987). 4: Afrotropic • Wingspan in metres (m): the total span of both wings from tip 5: Indo-Malay 6: Australasian to tip while fully extended. • • Wing area in square meters (m2): the total surface area of both Roost size: the largest reported roost size for the species, defin- wings including the intervening body area and area of the tail ing them by categories: membrane. 1: Solitary (1–3 bats) • Wing aspect ratio: calculated as the square of the wingspan 2: Small groups (4–20 bats) divided by wing area. 3: Moderate groups (21–99 bats) • Body mass (g): recorded from the individuals used to measure 4: Large groups (100–500 bats) wing parameters. 5: Highly gregarious (more than 500 bats) • 2 Wing loading (N/m ): calculated as the mass in kilograms (kg) Data analysis multiplied by g, the standard acceleration due to gravity ϳ The relationships between genome size and continuous pa- ( 9.80665 N/Kg) divided by the wing area. rameters were tested using Pearson correlations on log- Brain data transformed data. Where appropriate, body mass correction was employed by regression of both parameters against body

For personal use only. • Total brain mass in milligrams (mg). mass and subsequent Pearson correlations on the residuals. • Total brain volume in cubic millimeters (mm3). Categorical parameters (feeding categories, geographic distri- • Neocortex volume (mm3). bution, and roost size) were analyzed using ANCOVA with body • Body mass (g): recorded from the individuals that were used to mass as the covariable. Where the covariable is highly nonsig- measure brain weights and volumes. nificant (p > 0.20), ANOVA was completed. Significant differ- • Relative brain mass, and relative brain and neocortex volumes: ences between categories were identified using Tukey HSD calculated by dividing the values by body mass converted to tests. In addition, correlation analysis was performed on ge- milligrams (mg). nome size and principal component 1 (PC1), representing all • Relative neocortex volume to brain volume: calculated by di- body size parameters (body mass, head and body length, wing- viding the neocortex volume by the total brain volume. span, and wing area). Given the large number of correlations in this study, Bonferroni correction was also used to adjust the p Reproduction, development, and longevity data value to lower the risk of type I error. Given that many similar Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 • Gestation time: the length of uterine development in days. In parameters were used, correction was employed by counting species with delayed embryonic implantation or delayed devel- groups of parameters as correlations (e.g., Body size = body opment post implantation, only the time of continuous embry- mass, head and body length, etc.), giving a total of 10 grouped onic growth until parturition is recorded. comparisons. The significance level can be adjusted by Bonfer- • Birth weight: the weight in grams (g) of neonates; typically defined roni correction in several ways depending on how the dataset is as newly born young and occasionally near term embryos. considered to be partitioned. If considering all grouped param- • Time to weaning: the number of days that neonates are breast fed. eters at all taxonomic levels (species, genus, and family) and • Time to sexual maturity: typically the age of first reproduction in functional division (bats, megabats, and microbats) (n =10×3× both males and females; however, occasional data report the age 2 = 80), the correction sets p = 0.0006. Taken as independent at which functional sexual organs are present; values are given for datasets according to taxonomic level or functional division males and females separately. (n =10×3=30), p = 0.002; while considering just the grouped • Maximum recorded longevity for captive or wild individuals, in parameters (n = 10), p = 0.005. years. Accounting for phylogeny Ecological data To account for nonindependence of phylogenetically related species, analyses were completed using hierarchical taxonomic • Feeding categories: bats are divided categorically by diet: correlations (Pagel and Harvey 1988; Vinogradov 1995; Gregory 1: Insectivore 2000) using nested averages at the species, genus, and family 2: Frugivore/Nectarivore levels for all bats and microchiropterans, and at the species

Published by NRC Research Press Smith et al. 463

and genus level for megachiropterans. Additionally, though Fig. 1. (A) Summary of haploid genome size diversity in bats (black the current species-level phylogeny for Chiroptera is poorly bars; data from J.D.L. Smith and Gregory (2009) and the present resolved even at the genus level, phylogenetically independent study) relative to other mammalian species (grey bars; data from the contrasts (PICs) were attempted using the supertree presented Animal Genome Size Database; Gregory (2013)). Bats have a smaller by Jones et al. (2002) and Bininda-Emonds et al. (2007) for all average genome size relative to the mammalian average of ϳ3.5 pg significant or near significant relationships found using Pear- and show a distribution of genome sizes near the low end of the son correlations. PICs were conducted using the PDAP module mammalian dataset. (B) Summary of genome size diversity in (Midford et al. 2008) in Mesquite v2.5 (Maddison and Maddison 43 species of bats in the family Pteropodidae (megabats; grey bars; 2009), with one degree of freedom subtracted for each branch data from J.D.L. Smith and Gregory (2009)) and 121 species of in a polytomy and branch lengths set to 1. nonpteropodids (microbats) from 12 families (black bars; data from the present study). Results A Overview of genome size in bats 100 Bat Mammalian Average (3.50 pg) Genome size estimates for 121 species of microbats are pre- Average sented in Table 1. Of these, 94 represent novel genome size esti- (2.35 pg) mates, and 27 species have been previously studied (J.D.L. Smith 80 and Gregory 2009). Of those previously reported (most >20 years ago), our estimates were on average lower (2.40 pg vs. 2.54 pg; t test, p < 0.0008). 60 Genome size varied less than 2-fold in the species analyzed (including the megabats surveyed by J.D.L. Smith and Gregory 2009), ranging from 1.63 pg in Carriker’s round-eared bat, 40 Lophostoma carrikeri, to 3.17 pg in the lesser mouse-tailed bat, Rhinopoma hardwickii, with an average of 2.35 pg ± 0.02 SE. Thus, it ofNumber Species remained the case in this larger sample that all bats examined 20 possess genomes much smaller than those typical of most mam- mals (average ϳ3.50 pg) (t test, p < 0.0001; Fig. 1A)(Gregory 2013). The bat data were non-normally distributed around a mean of 0 2.35 pg ± 0.02 SE (Shapiro–Wilk test, W = 0.97, p < 0.05); however, 1234567 when analyzed separately the data for megabats were normally distributed around a mean of 2.20 pg ± 0.02 SE (Shapiro–Wilk test, Genome Size (pg) W = 0.98, p > 0.80), whereas the data for microbats were non- B normally distributed, averaging 2.40 ± 0.02 SE (Shapiro–Wilk test, 30 W = 0.97, p < 0.05). Megabats appear to be even more strongly Megabat Average Microbat Average constrained to small genome sizes than other bats in terms of (2.2 pg) (2.4 pg) both mean values (2.20 pg vs. 2.40 pg; t test, p < 0.0001; Fig. 1B) and 25

variance (F test, F[120,42] = 3.61, p < 0.0001) (J.D.L. Smith and Gregory For personal use only. 2009). 20 Chromosomal features As might be expected, chromosome number and fundamental Mammalian Average (3.5 pg) 15 number of arms were correlated within and across bat groups for genus and species level analyses (all r > 0.55; p < 0.0037), though nonsignificant at the family level. However, no significant rela- 10 tionship was found between genome size and chromosome num- of Species Number ber, fundamental number of arms, or the ratio between the two (all p > 0.32). 5

Cell size 0 Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 While sufficient cell size data were not available for statisti- 2.0 2.5 3.0 3.5 cal analyses within bats, it was possible to compare bats versus other mammals in terms of cell size. Figure 2 shows the rela- Genome Size (pg) tionship between genome size and erythrocyte size given as mean dry diameter and mean corpuscular volume in mammals, including the few bats for which these data were available. Bats highlighting the differences between morphology in pteropo- appear at the lower end of the plot with minimal deviation dids (“megabats”) and nonpteropodids (“microbats”) (but note from the main line, as would be expected if genome size and that the two do not form clear clusters as in some other ani- cell size do correlate in this group as among mammals in gen- mals; e.g., Ardila-Garcia and Gregory (2009)). Pearson correla- eral (Gregory 2001a). tions of genome size with PC1 (Fig. 3B) revealed no relationship for all bats (r = –0.1640, p = 0.2227, n = 56) or within microbats (r = Body size 0.0343, p = 0.8359, n = 39), or megabats (r = 0.4656, p = 0.096, n = 17). Body size can be measured using a variety of parameters, all Employing PICs weakened the relationship in megabats (r = of which are highly intercorrelated. Principal component anal- 0.3429, p = 0.1829, n = 16), whereas the relationship remained ysis was used on several morphometric parameters (body mass, insignificant in microbats (r = 0.2423, p = 0.1408, n = 38). head and body length, wingspan, and wing area) to produce a Body mass was found to correlate with genome size at the species single parameter (PC1) that accounted for ϳ96% of the variation level in both microbats (r = 0.2733, p = 0.0042, n = 108) and megabats in the combined dataset, while a second component (PC2) ac- (r = 0.4847, p = 0.0027, n = 36), but this was not so when all bats were counted for an additional 3%. Figure 3A plots PC1 versus PC2, analyzed together (r = –0.0341, p = 0.6846, n = 144). With PICs, the

Published by NRC Research Press 464 Genome Vol. 56, 2013

Fig. 2. Relationships between genome size and cell size in mammals, measured as (A) mean dry diameter (␮m) and (B) mean corpuscular volume (fL). Although data are not sufficient to examine correlations within bats, it is clear that their cell sizes fall along the same line (at the lowest end) as those of mammals in general. A

1.1

1.0

0.9

0.8

0.7

Other Mammals

Log Mean Dry Cell Diameter (µm) Diameter Dry Cell Log Mean 0.6 Bats Other Mammals All Mammals 0.5 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Log Genome Size (pg) B

2.2

2.0 For personal use only.

1.8

1.6

Other Mammals 1.4 Log Mean Corpuscular(fL) Volume Log Mean Bats

Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 Other Mammals All Mammals 1.2 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 Log Genome Size (pg)

relationship seen in megabats disappears (r = 0.0329, p = 0.8498, n = p = 0.9792, n = 14) and genus level for RBMR (uncorrected: r = –0.8141, 35), whereas it persists in microbats (r = 0.2947, p = 0.0024, n = 106). p = 0.0076, n = 9; mass-corrected: r = –0.5597, p = 0.1171, n = 9). No Head and body length correlated with genome size within correlations were found between genome size and body temperature megabats (species level; p = 0.0043; genus level, p = 0.0509); how- at any taxonomic level among all bats, within microbats, or within ever, after mass correction the relationship was not significant; megabats. PICs could not be performed on significant BMR correla- PICs revealed no relationship (r = 0.1910, p = 0.3535, n = 16). tions; data spread yielded contrasts that could not be correlated by Mesquite. Physiology Absolute (BMR) and relative (RBMR) basal metabolic rates did not Flight correlate with genome size at any level for all bats or within micro- No significant relationships were found between genome size and bats. Relationships between genome size and BMR and RBMR were wing parameters among all bats or within microbats at any taxo- only significant without mass correction at the species level for BMR nomic level. Wingspan, wing aspect ratio, and wing loading index (uncorrected: r = 0.632, p = 0.0153, n = 14; mass-corrected:r=–0.0077, were all correlated with body mass and highly intercorrelated (all

Published by NRC Research Press Smith et al. 465

Fig. 3. (A) Relationship between principal components 1 and 2 for all morphometric parameters, illustrating the separation and difference between pteropodids (megabats) and nonpteropodids (microbats). (B) Relationship between principal component 1 and genome size among all bats, within microbats and within megabats. A

1.5

1

0.5 PC 2

0

Microbats -0.5 Megabats Microbats Megabats -1 -2 0 2 4 6 8 10 PC 1

B

12 Microbats Megabats 10 All Bats For personal use only. Microbats 8 Megabats

6

4 PC 1

2

Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 0

-2

-4 1.80 2.00 2.20 2.40 2.60 2.80 3.00 3.20 Log Genome Size (pg)

r > 0.45, most p < 0.0001). Within megabats there was a significant Brain size correlation with wing loading index at the species level prior to mass Using multiple brain size parameters, no obvious relation- correction (uncorrected: r = 0.4507, p = 0.0461, n = 20; mass-cor- ships were found with genome size. Megabat genome size cor- rected: r = –0.2234, p = 0.2234, n = 19; Fig. 4). Prior to mass correc- related with all brain parameters prior to mass correction at tion, relationships between genome size and with wingspan, wing the species level (all p < 0.0172, all r > 0.49 for brain volume, area, and aspect ratio were marginal with all p < 0.1, with PICs; neocortex volume, BRBM and neocortex to brain volume and after mass correction these relationships were much weaker all r < –0.45 for relative brain mass, and relative brain and within megabats, showing no relationship. neocortex volumes); however, all relationships disappeared af-

Published by NRC Research Press 466 Genome Vol. 56, 2013

Fig. 4. Relationship between genome size and wing loading index within pteropodids (megabats; gray triangles, long-dashed lines), among nonpteropodids (microbats; black diamonds, short-dashed lines), and across all bats analyzed together (solid line), (A) before and (B) after mass correction. A 1.8 Microbats Megabats 1.6 Microbats )

2 Megabats

1.4 All Bats

1.2

1

0.8 Log Wing Loading Index (N/m

0.6

0.4 0.28 0.33 0.38 0.43 0.48 Log Genome Size (pg)

B

0.3 )

2 0.2

0.1 For personal use only.

0

-0.1

-0.2 Microbats Megabats

-0.3 Microbats Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 Megabats ResidualsWing Log Loading(N/m Index All Bats -0.4 -0.125 -0.075 -0.025 0.025 0.075 0.125 Residuals Log Genome Size (pg)

ter mass correction (all p > 0.15; Fig. 5); PICs yielded similar genome size based on biogeography revealed significant differ- results. All brain parameters were highly intercorrelated as ences between Indomalayan bats and Neotropical bats; however, might be expected (all r > 0.9, p < 0.0001). this reflects the differences in genome size between megabats and microbats. Reproduction, development, and longevity No relationships were found between genome size and any de- Discussion velopmental parameters (gestation time, birth weight, time to weaning, time to sexual maturity, and longevity). (1) What is the extent of genome size diversity in bats, and is this universally constrained relative to other mammals? Ecology Previous studies have indicated that genome size estimates No significant differences were found between genome size for bats are much lower on average than other mammalian and roost size or genome size and feeding categories. Analysis of groups. The suggested reason for this apparent genomic con-

Published by NRC Research Press Smith et al. 467

Fig. 5. Relationship between genome size and relative brain volume within pteropodids (megabats; gray triangles, long-dashed lines), among nonpteropodids (microbats; black diamonds, short-dashed lines), and across all bats analyzed together (solid line), at the species level, (A) before and (B) after mass correction. A -1.2

-1.4 /mg body 3

-1.6 mass) -1.8

Microbats

-2 Megabats Microbats Megabats Log RelativeVolume Brain (mm All Bats -2.2 0.275 0.325 0.375 0.425 0.475 0.525 Log Genome Size (pg)

B

0.3

0.2

0.1 For personal use only.

0

-0.1 /mg body mass) 3 -0.2 (mm

-0.3 Microbats Megabat

Residuals Log Relative Brain VolumeRelative Brain LogResiduals Microbats -0.4 Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 Megabats All Bats -0.5 -0.11 -0.06 -0.01 0.04 0.09 0.14 Residuals Log Genome Size (pg)

straint is the evolution of flight in bats, which precludes large overlap tends to occur in groups with high metabolic rates such as genomes (and cells) owing to the metabolic intensity of flight. shrews. This study greatly expanded the available dataset for bats, gen- Within bats, the megabats (family Pteropodidae) also have erating genome size estimates for 121 species from 12 families. small genomes (J.D.L. Smith and Gregory 2009), and, surprisingly, In agreement with previous work, all of these genome size they seem to have even more constrained genome sizes than mi- estimates are constrained relative to other mammals (Fig. 1). crobats in terms of both average and variance. While this may be Of the mammals studied to date, bats do indeed possess the expected since megabats comprise only one family whereas mi- smallest genomes. However, there is overlap with the low end of crobats represent many families, megabats have a lower average the distribution of some other orders. For example, some rodents genome size and equal or lower variance than other bat families display genome sizes as small as those at the higher end of the taken individually (Table 1). bat distribution. In keeping with the hypothesis that metabolic Genome size estimates within bat species tended to be consistent, constraints are relevant to patterns of genomic diversity, this usually varying less than 0.2 pg (i.e., within 8%), with standard errors

Published by NRC Research Press 468 Genome Vol. 56, 2013

less than ±0.05. In general, agreement with previous estimates pectation of a decoupling of genome size and body size is met. was good in terms of upholding similar interspecies differences. However, in more closely related groups with much less body size However, the new estimates were on average lower than esti- variability such as rodents, body size and genome size may be mates reported in the past for the same species. Differences in positively related (Gregory 2002c). current estimates and previous studies’ are likely due to differ- Principal component analysis to assess the relationship be- ences in methodologies (Hardie et al. 2002); previous estimates tween genome size and body size revealed a marginal relationship came from several different research groups, all of which used within megabats but not in microbats (except after employing differing methodologies (Capanna and Manfredi Romanini 1971; PICs). Unfortunately, multivariate techniques have the limitation Bachmann 1972a; Manfredi Romanini et al. 1975; Kato et al. 1980; of requiring data for every measure of body size assessed, result- Burton et al. 1989; Redi et al. 2005). ing in very small sample sizes in instances where only some pa- rameters are available for species. Using Pearson correlations (2) Are cytogenetic features related to genome size diversity revealed relationships between genome size and body mass in among bats? microbats and megabats and head and body length. Although the Chromosome number can be highly variable among animal relationships were not universal (with Bonferroni correction groups, in terms of both absolute number and the range of diver- some would be considered nonsignificant), they did not appear sity across related species. Chromosome numbers can be altered randomly. Rather, the relationships appeared at different taxo- by events such as deletions, duplications, inversions, and translo- nomic levels and both in microbats and megabats, not as would be cations, which can not only alter the overall structure of the chro- expected if the correlations were due to chance when completing mosome complement but may in some cases result in a change in large numbers of correlations. Although not strongly related, DNA content. Furthermore, given that necessary structural com- measures of body size do correlate with genome size, particularly ponents such as centromeres and telomeres are composed of re- in megabats where body size differences are greater and detect- petitive DNA, it might be expected that some of this variability able relationships are stronger. With PICs similar results were would be related to total DNA content. These factors may account found, although the relationship between genome size and body for the reported correlation between genome size and diploid size in megabats may be related to phylogeny. Unfortunately, the chromosome number in ray-finned fishes (Mank and Avise 2006; best tree available for bats is still unresolved at the species level E.M. Smith and Gregory 2009). for many taxa, with a large number of polytomies (ϳ35% of the As with genome size, bats have somewhat lower diploid chro- tree) making it difficult to draw any firm conclusions. mosome numbers (average: 2n ≈ 38, range: 2n = 16 to 62) than the The relationship between genome size and body size could be mammalian average (2n ≈ 44) (Neuweiler 2000; Ruvinsky and explained by changes in cell volume, as has been found in some Marshall Graves 2004). However, within bats, no relationship was invertebrates. However, it is also possible that the relationship is found between genome size and diploid chromosome number, somewhat spurious and results from links of both genome size fundamental number of chromosome arms, or the ratio between and body size to an additional trait. Body size is related to a large the two parameters, suggesting that chromosome level altera- number of potentially adaptive biological features. This relation- tions are not responsible for overall genome size variation within ship makes it somewhat difficult to determine which traits are this order. acting adaptively to modulate genome size because mass correc- While chromosome-level mechanisms do not seem to play a tion removes so much variation that often no residual relation- major role in determining genome size diversity among bats, ships can be found. While adaptive mechanisms may lead to a

For personal use only. there are other genomic features that may be relevant. Transpos- relationship between genome size and body size, it is also possible able elements (TEs) in particular represent a substantial portion of that nonadaptive mechanisms are involved. Lynch and Conery mammalian genomes—45% of the human genome is comprised (2003) proposed that genome size can evolve through accumula- of TEs, for example (International Human Genome Sequencing tion of slightly deleterious duplications and TE insertions fixed by Consortium 2001). Differences in TE composition between species genetic drift in small populations. Body mass has been used as an could explain a large amount of the genome size variation. Nota- estimator of population size and relationships between genome bly, megabats experienced an extinction of the most common size and body size may simply reflect passive accumulation of long interspersed nuclear element in mammals (LINE-1) early in DNA (Lynch 2007). This study did examine roost size in bats, find- their ancestry, which may help to explain the reduction in abso- ing no relationship between genome size and population size; lute numbers of these elements and a correlated decrease of ge- however, whether roost size can be taken as a measure of effective nome size in the megabats (Cantrell et al. 2008). This may be population size is uncertain, leaving this possibility open. The additionally relevant because short interspersed nuclear ele- link between genome size and body size in bats (and rodents)

Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 ments (SINEs) are dependent on LINEs for their mobility. On the requires further investigation to tease apart the underlying mech- other hand, recent evidence in the little brown bat (Myotis lucifu- anisms. gus) has suggested that while most TEs are thought to be inactive in mammals, there has been more recent activity in some species (4) Can constraints related to flight explain the observed of bats (Ray et al. 2007, 2008). This is similar to the situation found diversity of genome size among bats? in pufferfishes, in which TEs are not abundant but are diverse and Flight is thought to be a constraining factor on genome size in active, perhaps reflecting divergence of elements under strict birds and bats through a link between genome size, cell size, and competition imposed by limited insertion sites in small genomes metabolic rate. Flight is a metabolically expensive activity leading (Neafsey and Palumbi 2003; Gregory 2005b). to very high mass-specific metabolisms in bats. Genome size and cell size positively correlate within mammals (Gregory 2000); (3) Are differences in genome size among bats linked to however, in bats, too few cell size data were available to examine body size? the link between genome size and cell size (and hence metabo- Initial hypotheses relating genome size to body size stem from lism). Nonetheless, the few data that were available fit at the lower invertebrate groups such as copepods and flatworms whose body end of the relationship found between cell size and genome size size is determined by changes in cell volume, rather than changes in other mammals, as would be expected (Fig. 2). in cell number (Gregory et al. 2000). Such relationships are not Previous studies have found that in both mammals and birds typically expected in groups like mammals, where genome size there is a relationship between mass specific metabolic rate and ranges 4-fold but body sizes differ by several orders of magnitude. genome size (Vinogradov 1995, 1997; Gregory 2002a). Relationships While examining mammals at higher taxonomic levels, this ex- were seen in megabats between genome size and absolute and

Published by NRC Research Press Smith et al. 469

Fig. 6. Relationship between genome size and relative metabolic rate within pteropodids (megabats; gray triangles, long-dashed lines), among nonpteropodids (microbats; black diamonds, short-dashed lines), and across all bats analyzed together (solid line), (A) before and (B) after mass correction. A 0.6 Microbats 0.5 Megabats Microbats 0.4 Megabats All Bats 0.3 /g/hr) 2

0.2

0.1

0

-0.1

-0.2 Log(mL.O MetabolicRate

-0.3

-0.4 0.28 0.33 0.38 0.43 0.48 Log Genome Size (pg)

B

0.3

0.2 /g/hr) 2 For personal use only.

0.1

0

-0.1

-0.2 Microbats Megabats Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 -0.3 Microbats

Residuals Log Metabolic Rate (mL.O Metabolic Residuals Log Megabats All Bats -0.4 -0.1 -0.05 0 0.05 0.1 0.15 Residuals Log Genome Size (pg)

relative basal metabolic rates but not in microbats. This might be be sufficiently influential to determine the narrow range of ge- expected as absolute and relative metabolic rates scale with body nome size in modern bats. The second possibility is that BMR does mass, and megabats have a much greater variation in body size still relate to genome sizes among bats but that a relationship (White and Seymour 2003). However, absolute basal metabolism was not found for a number of reasons including the following: and relative basal metabolic rate in bats were not correlated with (i) Measurements of basal metabolism are particularly sensitive to genome size in bats when the influence of body mass was re- variation in experimental technique, but they were necessarily moved statistically (Fig. 6). sourced from several different references. It cannot be guaranteed The lack of relationship between genome size and metabolism that cross comparison is accurate as this level of error may inter- can be interpreted in many ways. The first possibility would be fere with detection of relationships with such a narrow range in that while a high metabolism may have been important in an genome size. (ii) Considering basal metabolism as the informative early reduction in genome size in the ancestors of bats, it may not metabolic measurement, rather than active metabolism or meta-

Published by NRC Research Press 470 Genome Vol. 56, 2013

bolic scope (for which data are unavailable) may not be the most velopment owing to their volant lifestyle and the burden of car- informative. (iii) Bats have unique features that confound studies rying young. on energetics; for example, they are capable of long-term hiber- nation and daily torpor to reduce energy usage. These unique (7) Is diversity in genome size among bats associated with patterns of energy distribution make it difficult to interpret single any ecological features? measures of metabolism. (iv) The relationship between genome Many of the reasons for expecting genome size to correlate with size and body mass may remove so much variation that after mass ecological features are dependent on relationships with metabo- correction there is insufficient variation to yield a significant re- lism, cognition, and development. These characters do not appear lationship with current sample sizes. A third interpretation re- to be related to genome size in bats, so it may not be surprising lates to the fact that, like all mammals, bats have enucleated that roost size, diet, and biogeography were unrelated to genome erythrocytes, which allows for more compact cells compared with size in this study. other vertebrate groups with similar genome sizes. Erythrocyte size still correlates with genome size in mammals (Gregory 2000); Conclusions however, it is possible that this mechanism of achieving small Based on the results of the present research, the following three cells and high metabolism allows for less stringent selection on major conclusions can be drawn regarding genome size diversity genome size than is seen for example in nucleated erythrocytes in in bats: birds. While bats have high metabolisms and low genome size, simi- (1) Genome size in both bat groups is constrained relative to lar to birds, they differ as no correlation exists between metabo- other mammals as expected; however, while they might both lism and genome size among bats. Notably, parameters that be expected to be similarly constrained on the basis of flight, indicate specialization for flight such as wing loading index are there are some interesting differences between microbats related to genome size in birds but not in bats (Andrews et al. and megabats illustrated by the relationships of genome size 2009). So while bats have low genome size relative to other mam- with various parameters. mals, suggesting that small genomes may in some way be associ- (2) Genome size in bats does not appear to be related to most ated with flight, current levels of variation in genome size among biological parameters that have been found to correlate in bats do not seem to be due to adaptations for flight. Again this other vertebrate taxa, with the exception of a relationship could be related to novel innovations of erythrocytes, which may with body size. While it is possible that genome size relates to decouple the link between genome size, cell size, and metabolism body size owing to changes in cell volume, it is also possible that can be seen in birds. Another possibility is that small ge- (and perhaps more likely) that genome size is sculpted by nomes and high metabolic rates evolved together early in the bat adaptive parameters that correlate with body size. It is also lineage and have been maintained since, leading to minimal vari- possible that this reflects the role of nonadaptive processes ation that can correlate strongly among modern bats. relating to population size. The significance of the body size correlation, therefore, remains an open question. (5) Are neurological constraints relevant to genome size (3) The hypothesis that flight was a constraining factor on ge- diversity among bats? nome size early in bat evolution is strongly supported, but Brain size in bats has been found to relate to flexibility in feed- variability in flight intensity does not explain the small dif- ing behaviours, whereby bats with larger brains have the ability to ferences in DNA amount observed in modern bat species.

For personal use only. utilize more dispersed feeding areas, can occupy more complex habitats, and can feed using different mechanisms (Eisenberg and Future directions Wilson 1978; Safi and Dechmann 2005; Ratcliffe et al. 2006). Sim- Although the present study has contributed significantly to in- ilarly, in birds larger brains appear to increase survival in novel formation on bat genome size, many interesting topics remain to environments (Sol et al. 2005). Recently, relative brain size was be explored in this group. Some of these are outlined as follows: found to negatively correlate with genome size in parrots, indi- (1) While the fossil record is limited for bat species, it might be cating a higher relative investment in brain tissue and (or) com- worthwhile to estimate genome size from extinct species. plexity (Andrews and Gregory 2009). Unlike in birds, relative brain size does not correlate with genome size in bats independent of Transitional forms would be ideal to test the hypothesis that body mass. However, relative brain volume does correlate with genome size must be small for flight to evolve; however, even body mass, suggesting that brain size is increased to improve extinct flighted species could be informative. If bats follow brain function or give equivalent brain power to larger animals, the patterns seen in dinosaurs/birds and pterosaurs, then

Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 as would be expected, accomplished by increasing brain size, not there should be signs of genome size decrease early in bat by decreasing cell size. evolution. (2) Transitional forms of bats are predicted to be similar to mod- (6) Are developmental parameters related to diversity in ern day colugos or flying squirrels, with membranes genome size among bats? stretched between limbs for gliding. Studies concerning the Developmental rate is often thought to correlate with genome relative metabolisms and genome size of gliding mammals size because of corresponding influences of DNA amount on cell and their closest nongliding relatives would perhaps illus- cycle duration. Relationships between genome size and develop- trate the importance of small genomes for prevolant life- mental rate and intensity of metamorphosis have been found styles. within amphibians (Gregory 2002b); however, in mammals, a (3) The relationship between genome size and cell size has been developmental relationship is only apparent within rodents examined almost entirely on the basis of erythrocytes among (Gregory 2002c) and was not observed in primates (Morand and vertebrates. It would be helpful to examine other cell types in Ricklefs 2005). mammals and other taxa. This could be especially informa- Based on the present analysis, genome size does not correlate tive for determining the cause of the relationship between with any developmental parameters or longevity in bats. As ge- genome size and body size in bats. nome size in bats is quite constrained, it is possible that small (4) The notion that passive accumulation of DNA in small popu- changes in genome size may not be enough to cause appreciable lations is a determinant of diversity in genome size remains differences in cell cycle length. However, it is probable that bats to be tested for most groups. Reliable genetic methods for are uniquely constrained in terms of their reproduction and de- estimating long-term effective population size could be use-

Published by NRC Research Press Smith et al. 471

ful, especially in groups such as bats where population sizes Gregory, T.R. (Editor). 2005a. Genome size evolution in animals. In The evolution vary widely. of the genome. Elsevier. pp. 3–87. Gregory, T.R. 2005b. Synergy between sequence and size in large-scale genomics. Bats have highly constrained genomes relative to other mam- Nat. Rev. Genet. 6: 699–708. doi:10.1038/nrg1674. PMID:16151375. Gregory, T.R. 2013. Animal genome size database. Available from http:// mals, overlapping in genome size with many avian species. While www.genomesize.com [accessed 19 September 2013]. clear comparisons can be drawn between birds and bats in terms Gregory, T.R., and Witt, J.D.S. 2008. Population size and genome size in fishes: a of their genomes, the biological impact of this diversity may differ closer look. Genome, 51(4): 309–313. doi:10.1139/G08-003. PMID:18356967. between the two groups. Thus, while flight may have constrained Gregory, T.R., Hebert, P.D.N., and Kolasa, J. 2000. Evolutionary implications of genome size in the early ancestry of all three lineages of flying the relationship between genome size and body size in flatworms and cope- pods. Heredity, 84: 201–208. doi:10.1046/j.1365-2540.2000.00661.x. PMID: vertebrates (bats, birds, and pterosaurs), flight-related effects are 10762390. not important in terms of patterns within bats in the manner that Grosset, L., and Odartchenko, N. 1975a. Duration of mitosis and separate mitotic they are in birds. Instead, other factors (particularly those directly phases compared to nuclear DNA content in erythroblasts of four verte- or indirectly linked to body size) represent the most significant brates. Cell Tissue Kinet. 8: 91–96. doi:10.1111/j.1365-2184.1975.tb01210.x. PMID:1089478. areas for further exploration in airborne mammals. Grosset, L., and Odartchenko, N. 1975b. Relationships between cell cycle dura- tion, S-period and nuclear DNA content in erythroblasts of four vertebrate Acknowledgements species. Cell Tissue Kinet. 8: 81–90. doi:10.1111/j.1365-2184.1975.tb01209.x. This research was supported by a Natural Sciences and Engi- PMID:1089477. neering Research Council of Canada (NSERC) post-graduate schol- Hardie, D.C., Gregory, T.R., and Hebert, P.D.N. 2002. From pixels to pico- grams: a beginners’ guide to genome quantification by Feulgen image arship to J.D.L.S. and an NSERC grant to T.R.G. The authors wish to analysis densitometry. J. Histochem. Cytochem. 50: 735–749. doi:10.1177/ thank John Patton at Purdue University, the staff of the Lubee Bat 002215540205000601. PMID:12019291. Conservancy, Burton Lim and colleagues at the Royal Ontario Hughes, A.L. 1999. Adaptive evolution of genes and genomes. Oxford University Museum, and the staff at the Museum of the North for facilitating Press, New York. Hughes, A.L., and Hughes, M.K. 1995. Small genomes for better flyers. Nature, access to the samples that made this study possible. The authors 377: 391. doi:10.1038/377391a0. PMID:7566113. also thank Jillian Bainard and the reviewers for helpful comments International Human Genome Sequencing Consortium. 2001. Initial sequencing that improved the paper. and analysis of the human genome. Nature, 409: 860–921. doi:10.1038/ 35057062. PMID:11237011. References Jones, K.E., Purvis, A., MacLarnon, A., Bininda-Emonds, O.R.P., and Simmons, N.B. 2002. A phylogenetic supertree of the bats (Mammalia: Chirop- Andrews, C.B., and Gregory, T.R. 2009. Genome size is inversely correlated with tera). Biol. Rev. Camb. Philos. Soc. 77: 223–259. doi:10.1017/S1464793101005899. relative brain size in parrots and cockatoos. Genome, 52(3): 261–267. doi:10. PMID:12056748. 1139/G09-003. PMID:19234554. Kato, H., Harada, M., Tsuchiya, K., and Moriwaki, K. 1980. Absence of correlation Andrews, C.B., Mackenzie, S.A., and Gregory, T.R. 2009. Genome size and wing between DNA repair in ultraviolet irradiated mammalian cells and life span parameters in passerine birds. Proc. R. Soc. B Biol. Sci. 276: 55–61. doi:10.1098/ of the donor species. Jpn. J. Genet. 55: 99–108. doi:10.1266/jjg.55.99. rspb.2008.1012. Li, G., Wang, J., Rossiter, S.J., Jones, G., Cotton, J.A., and Zhang, S. 2008. The Ardila-Garcia, A.M., and Gregory, T.R. 2009. An exploration of genome size di- hearing gene Prestin reunites echolocating bats. Proc. Natl. Acad. Sci. U.S.A. versity in dragonflies and damselflies (Insecta: Odonata). J. Zool. (Lond.), 278: 105: 13959–13964. doi:10.1073/pnas.0802097105. PMID:18776049. 163–173. doi:10.1111/j.1469-7998.2009.00557.x. Lynch, M. 2007. The origins of genome architecture. Sinauer Associates, Sunder- Bachmann, K. 1972a. Genome size in mammals. Chromosoma, 37: 85–93. doi: land, Mass. 10.1007/BF00329560. PMID:5032813. Lynch, M., and Conery, J.S. 2003. The origins of genome complexity. Science, Bachmann, K. 1972b. Nuclear DNA and developmental rate in frogs. Quart. J. Fla. 302: 1401–1404. doi:10.1126/science.1089370. PMID:14631042. Acad. Sci. 35: 225–231.

For personal use only. Maddison, W.P., and Maddison, D.R. 2009. Mesquite: A modular system for Bailey, W.J., Slightom, J.L., and Goodman, M. 1992. Rejection of the “flying pri- evolutionary analysis, Version 2.6. [Online.] Available from http:// mate” hypothesis by phylogenetic evidence from the ␧-globin gene. Science, mesquiteproject.org. 256: 86–89. doi:10.1126/science.1301735. PMID:1301735. Manfredi Romanini, M.G., Pelliciari, C., Bolchi, F., and Capanna, E. 1975. Don- Bininda-Emonds, O.R.P., Cardillo, M., Jones, K.E., MacPhee, R.D.E., Beck, R.M.D., nées nouvelles sur le contenu en ADN des noyaux postkinétiques chez les Grenyer, R., et al. 2007. The delayed rise of present-day mammals. Nature, chiroptères. Mammalia, 39: 675–683. PMID:1225677. 446: 507–512. doi:10.1038/nature05634. PMID:17392779. Mank, J.E., and Avise, J.C. 2006. Phylogenetic conservation of chromosome num- Burton, D.W., Bickham, J.W., and Genoways, H.H. 1989. Flow-cytometric analy- bers in Actinopterygiian fishes. Genetica, 127: 321–327. doi:10.1007/s10709- ses of nuclear DNA content in four families of neotropical bats. Evolution, 43: 005-5248-0. PMID:16850236. 756–765. doi:10.2307/2409304. Midford, P.E., Garland, T., and Maddison, W.P. 2008. PDAP:PDTREE Package for Cantrell, M.A., Scott, L., Brown, C.J., Martinez, A.R., and Wichman, H.A. 2008. Mesquite, Version 1.14. [Online.] Available from http://mesquiteproject.org/ Loss of LINE-1 activity in the megabats. Genetics, 178: 393–404. doi:10.1534/ pdap_mesquite/. genetics.107.080275. PMID:18202382. Morand, S., and Ricklefs, R.E. 2001. Genome size, longevity and development Capanna, E., and Manfredi Romanini, M.G. 1971. Nuclear DNA content and mor- time in birds. Trends Genet. 17: 567–568. doi:10.1016/S0168-9525(01)02414-3. phology of the karyotype in certain palearctic Microchiroptera. Caryologia, PMID:11642250.

Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16 24: 471–482. Morand, S., and Ricklefs, R.E. 2005. Genome size is not related to life-history Eisenberg, J.F., and Wilson, D.E. 1978. Relative brain size and feeding strategies traits in primates. Genome, 48(2): 273–278. doi:10.1139/g04-125. PMID: in the Chiroptera. Evolution, 32: 740–751. doi:10.2307/2407489. 15838549. Gregory, T.R. 2000. Nucleotypic effects without nuclei: genome size and eryth- Neafsey, D.E., and Palumbi, S.R. 2003. Genome size evolution in pufferfish: a rocyte size in mammals. Genome, 43(5): 895–901. doi:10.1139/g00-069. PMID: comparative analysis of Diodontid and Tetraodontid pufferfish genomes. 11081981. Genome Res. 13: 821–830. doi:10.1101/gr.841703. PMID:12727902. Gregory, T.R. 2001a. The bigger the C-value, the larger the cell: genome size and Neuweiler, G. 2000. The biology of bats. Oxford University Press, New York. red blood cell size in vertebrates. Blood Cells Mol. Dis. 27: 830–843. doi:10. Norberg, U.M., and Rayner, J.M.V. 1987. Ecological morphology and flight in bats 1006/bcmd.2001.0457. PMID:11783946. (Mammalia; Chiroptera): wing adaptations, flight performance, foraging Gregory, T.R. 2001b. Coincidence, coevolution, or causation? DNA content, cell strategy and echolocation. Philos. Trans. R. Soc. B Biol. Sci. 316: 335–427. size, and the C-value enigma. Biol. Rev. Camb. Philos. Soc. 76: 65–101. doi:10. doi:10.1098/rstb.1987.0030. 1111/j.1469-185X.2000.tb00059.x. PMID:11325054. Organ, C.L., and Shedlock, A.M. 2009. Paleogenomics of pterosaurs and the Gregory, T.R. 2002a. A bird’s-eye view of the C-value enigma: genome size, cell evolution of small genomesize in flying vertebrates. Biol. Lett. 5: 47–50. size, and metabolic rate in the class Aves. Evolution, 56: 121–130. doi:10.1111/ doi:10.1098/rsbl.2008.0491. PMID:18940771. j.0014-3820.2002.tb00854.x. PMID:11913657. Organ, C.L., Shedlock, A.M., Meade, A., Pagel, M., and Edwards, S.V. 2007. Origin Gregory, T.R. 2002b. Genome size and developmental complexity. Genetica, 115: of avian genome size and structure in non-avian dinosaurs. Nature, 446: 131–146. doi:10.1023/A:1016032400147. PMID:12188045. 180–184. doi:10.1038/nature05621. PMID:17344851. Gregory, T.R. 2002c. Genome size and developmental parameters in the homeo- Pagel, M.D., and Harvey, P.H. 1988. Recent developments in the analysis of thermic vertebrates. Genome, 45(5): 833–838. doi:10.1139/g02-050. PMID: comparative data. Q. Rev. Biol. 63: 413–440. doi:10.1086/416027. PMID: 12416615. 3068704. Gregory, T.R. 2003. Variation across amphibian species in the size of the nuclear Pettigrew, J.D. 1986. Flying primates? Megabats have the advanced pathway genome supports a pluralistic, hierarchical approach to the C-value enigma. from eye to midbrain. Science, 231: 1304–1346. PMID:3945827. Biol. J. Linn. Soc. Lond. 79: 329–339. doi:10.1046/j.1095-8312.2003.00191.x. Ratcliffe, J.M., Fenton, M.B., and Shettleworth, S.J. 2006. Behavioral flexibility

Published by NRC Research Press 472 Genome Vol. 56, 2013

positively correlated with relative brain volume in predatory bats. Brain Smith, F.A., Lyons, S.K., Ernest, S.K.M., Jones, K.E., Kaufman, D.M., Dayan, T., Behav. Evol. 67: 165–176. doi:10.1159/000090980. PMID:16415571. et al. 2003. Body mass of late quaternary mammals. Ecology, 84: 3403. doi: Ray, D.A., Pagan, H.J.T., Thompson, M.L., and Stevens, R.D. 2007. Bats with hATs: 10.1890/02-9003. evidence for recent DNA transposon activity in genus Myotis. Mol. Biol. Evol. Sol, D., Duncan, R.P., Blackburn, T.M., Cassey, P., and Lefebvre, L. 2005. Big 24: 632–639. doi:10.1093/molbev/msl192. PMID:17150974. brains, enhanced cognition, and response of birds to novel environments. Ray, D.A., Feschotte, C., Pagan, H.J.T., Smith, J.D., Pritham, E.J., Arensburger, P., Proc. Natl. Acad. Sci. U.S.A. 102: 5460–5465. doi:10.1073/pnas.0408145102. et al. 2008. Multiple waves of recent DNA transposon activity in the bat, PMID:15784743. Myotis lucifugus. Genome Res. 187: 17–28. doi:10.1101/gr.071886.107. Szarski, H. 1983. Cell size and the concept of wasteful and frugal evolutionary Redi, C.A., Zacharias, H., Merani, S., Oliveira-Miranda, M., Aguilera, M., strategies. J. Theor. Biol. 105: 201–209. doi:10.1016/S0022-5193(83)80002-2. Zuccotti, M., et al. 2005. Genome sizes in Afrotheria, Xenarthra, Euarchontoglires, PMID:6656279. and Laurasiatheria. J. Hered. 96: 485–493. doi:10.1093/jhered/esi080. PMID: Teeling, E.C., Madsen, O., Van Den Bussche, R.A., de Jong, W.W., Stanhope, M.J., 15994420. and Springer, M.S. 2002. Microbat paraphyly and the convergent evolution of Roth, G., Blanke, J., and Wake, D.B. 1994. Cell size predicts morphological com- a key innovation in Old World rhinolophoid microbats. Proc. Natl. Acad. Sci. plexity in the brains of frogs and salamanders. Proc. Natl. Acad. Sci. U.S.A. 91: U.S.A. 99: 1431–1436. doi:10.1073/pnas.022477199. PMID:11805285. 4796–4800. doi:10.1073/pnas.91.11.4796. PMID:8197137. Van’t Hof, J., and Sparrow, A.H. 1963. A relationship between DNA content, Ruvinsky, A., and Marshall Graves, J.A. 2004. Mammalian genomics. CABI Pub- nuclear volume, and minimum mitotic cycle time. Proc. Natl. Acad. Sci. lishing, Cambridge, UK. U.S.A. 49: 897–902. doi:10.1073/pnas.49.6.897. PMID:13996145. Safi, K., and Dechmann, D.K.N. 2005. Adaptation of brain regions to habitat Vinogradov, A.E. 1995. Nucleotypic effect in homeotherms: body-mass-corrected complexity: a comparative analysis in bats (Chiroptera). Proc. R. Soc. B Biol. basal metabolic rate of mammals is related to genome size. Evolution, 49: Sci. 272: 179–186. doi:10.1098/rspb.2004.2924. 1249–1259. doi:10.2307/2410449. Smith, E.M., and Gregory, T.R. 2009. Patterns of genome size diversity in the Vinogradov, A.E. 1997. Nucleotypic effect in homeotherms: body-mass indepen- ray-finned fishes. Hydrobiologia, 625: 1–25. doi:10.1007/s10750-009-9724-x. dent metabolic rate of passerine birds is related to genome size. Evolution, Smith, J.D.L. 2009. Evolution of genome size in bats (mammalia: chiroptera). 51: 220–225. doi:10.2307/2410975. M.Sc. thesis, Department of Integrative Biology, University of Guelph, White, C.R., and Seymour, R.S. 2003. Mammalian basal metabolic rate is propor- Guelph, Ontario, Canada. tional to body mass2/3. Proc. Natl. Acad. Sci. U.S.A. 100: 4046–4049. doi:10. Smith, J.D.L., and Gregory, T.R. 2009. The genome sizes of megabats (Chiroptera: 1073/pnas.0436428100. PMID:12637681. Pteropodidae) are remarkably constrained. Biol. Lett. 5: 347–351. doi:10.1098/ Wilson, D.E., and Reeder, D.M. 2005. Mammal species of the world. Johns Hop- rsbl.2009.0016. PMID:19324635. kins University Press, Baltimore. For personal use only. Genome Downloaded from www.nrcresearchpress.com by 65.74.4.35 on 01/25/16

Published by NRC Research Press